The Semantic Data Charter (SDC) is an open specification for creating self-describing, semantically rich data models that work seamlessly across systems, languages, and autonomous agents.

This organization and all repositories within it are controlled and maintained by Axius SDC, Inc.

Data infrastructure engineered for fully automated machine processing.

We saw where technology was heading 25 years ago: machines would become the primary consumers and producers of structured data. When that happened, every system built for human interpretation would break. SDC was built for that world.

• Open - Built on international standards, not proprietary lock-in
• Accessible - Language-agnostic, enabling global collaboration
• Governed - Built-in lineage and provenance for accountability
• Permanent - CUID-based immutability for long-term data integrity

1. Enforce Governance - Built-in lineage and provenance tracking
2. Embed Meaning - Semantic interoperability via RDF/OWL
3. Mandate Quality - Validation rules and quality constraints

Think of it as: "Protocol Buffers with semantics and governance"

SDC combines:

• Structure (like Protocol Buffers) - XML Schema 1.1 for data types and constraints
• Semantics (like RDF/OWL) - Meaning and relationships
• Validation (like SHACL) - Quality rules and business logic
• Governance - Lineage, provenance, and audit trails
• Packaging - Self-contained, queryable data packages

Result: Data that is self-describing, semantically rich, and audit-ready - readable by machines without human interpretation.

• Language Agnostic - Model in Portuguese, French, Japanese, Spanish, etc.
• Standards-Aligned - Aligned with the vocabularies of 20+ international standards from eight bodies (W3C, ISO, OASIS, OMG, IETF, IEEE, IANA, NIST), enumerated in SDCRM/STANDARDS.md
• AI Governance - Governance decision engine returning OASIS XACML decisions with tamper-evident receipts
• Namespace Versioning - CUIDs for immutable, evolvable components
• Multi-Format Export - RDF, OWL, SHACL, FHIR, GraphQL, JSON, GQL

• Enterprise Data Integration - Canonical models for multi-system environments
• AI Governance - Data packages with automatic lineage for trustworthy AI
• Cross-Border Exchange - Semantic interoperability without translation
• Regulatory Compliance - Built-in audit trails and provenance
• Legacy Modernization - Bridge between old and new systems
• Agentic Infrastructure - Deterministic data substrate for autonomous agents

• SDCRM v4.1.0 - Reference model, schemas, and specification (releases)
• sdc-xml2graph v4.0.0 - Transform SDC4 XML to knowledge graphs (Q1-2026)

• Form2SDCTemplate v4.4.0 - Convert PDF, DOCX, and image forms into SDC templates using Gemini AI. Available as a Python package, Google Colab notebook, or LLM instruction file.
• SDCObsidianTemplate v4.3.0 - Interactive Obsidian Templater plugin for building SDC markdown templates with guided prompts, domain-aware defaults, and SDC participation model support.

• sdcvalidator (Python) v4.5.2 - SDC4 structural validator with two-tier error classification (PyPI)
• sdcgovernance (Python) v4.2.3 - Governance decision engine, aligned with the vocabularies of 15 formal standards, nine of them core (per-repository inventory in SDCRM/STANDARDS.md). Returns OASIS XACML decisions (PERMIT/DENY/INDETERMINATE/NOT_APPLICABLE) with SHA-256 hash-chained tamper-evident receipts. MCP server included. (PyPI)
• sdcreceipt (Python) v4.2.2 - Verify and settle VSL Settlement Receipts. Offline verification against published keys, no account and no network; party key setup, trigger signing, settle; eleven published conformance vectors; MCP server in the registry as io.github.SemanticDataCharter/sdcreceipt. (PyPI)

• SDCStudio v4.0.0 - Web application for AI-powered SDC model generation, schema export, and semantic enrichment

All SDC projects use 4.x.x versioning where the MAJOR version (4) represents the current SDC generation.

You will see work under two GitHub organizations. The split is deliberate and the axis is stewardship, not licensing.

• SemanticDataCharter holds the specification and its open reference implementation: the reference model, the open component library, and the tools that validate and govern against it. This is the part intended to outlive any single company, and it is the part a foundation home would take. Everything here is Apache License 2.0. A repository that is temporarily private is unreleased, not proprietary.
• Axius-SDC holds the work of Axius SDC, Inc., the company that maintains SDC: products, demos and commercial platforms. Some of it is open source too, for example CordovaOS. Licensing there is per repository; check the LICENSE file.

The short version: this organization is the standard, the other is the company. Contributions to repositories here are covered by the Contributor License Agreement.

AxiusSDC, without the hyphen, is not an organization we use. If you find a link pointing at it, it is a typo and we would be glad to hear about it.

SDC was not built in response to the AI era. It was built in anticipation of it.

Evolution Timeline:

• 2000-2009: FreePM/TORCH (open source healthcare applications)
• 2009-2017: MLHIM (430 commits, 160+ citations, healthcare focus)
• 2012-2025: S3Model (1,586 commits, domain-agnostic generalization)
• 2025-present: SDC4 (commercial production platform)

Lineage Validation (MLHIM and S3Model):

• 10 peer-reviewed publications, 2002 to 2016 (Elsevier, Springer, IEEE, BMC), plus a JAMA letter; full list with DOIs
• 160+ citations on Google Scholar
• Applied in healthcare, research, and enterprise contexts

These attach to the predecessor projects. SDC4 inherits the modeling approach, not the citation record.

Provenance: The full development history from 2013 is preserved in a private archive, available for due diligence on request.

SDC aligns with the vocabularies of international standards:

W3C Standards:

• XML Schema 1.1 (structure)
• RDF 1.1 (semantics)
• OWL 2 (ontologies)
• SPARQL 1.1 (queries)
• SHACL (validation)

ISO Standards:

• ISO/IEC 21838 (top-level ontologies)
• ISO 21090 (healthcare data types)

IETF Standards:

• RFC 3986 (URIs)
• RFC 8259 (JSON)
• And more...

OASIS Standards:

• XACML 3.0 (governance decisions)

OMG Standards:

• DMN (decision tables)

Total: 20+ international standards from eight bodies (W3C, ISO, OASIS, OMG, IETF, IEEE, IANA, NIST), enumerated in SDCRM/STANDARDS.md. That file is the authoritative inventory; the lists above are examples.

The fastest path from existing form to SDC model:

1. Open the Form2SDCTemplate Colab notebook
2. Enter your free Google AI API key
3. Upload a PDF, DOCX, or image form
4. Download the generated SDC markdown template
5. Upload to SDCStudio for processing

Or use the Python package:

pip install "form2sdc[gemini]"

from form2sdc.analyzer import GeminiAnalyzer
from form2sdc.core import FormToTemplatePipeline
from pathlib import Path

analyzer = GeminiAnalyzer(api_key="YOUR_KEY")
pipeline = FormToTemplatePipeline(analyzer)
result = pipeline.process(Path("your_form.pdf"))

print(result.template)          # SDC markdown
print(result.validation.valid)  # True if valid

pip install sdcvalidator

from sdcvalidator import SDC4Validator

validator = SDC4Validator('schema.xsd')
result = validator.validate('data.xml')

if result.is_valid:
    print("Valid!")
else:
    print(f"Structural errors: {len(result.structural_errors)}")
    print(f"Semantic errors: {len(result.semantic_errors)}")

pip install sdcgovernance

from sdcgovernance import validate_governance, GovernanceEngine

# Primary API: validate governance content in an instance against its model
result = validate_governance("model.xsd", "instance.xml")
print(result.decision)       # Decision.PERMIT

# Stateful engine, for agents evaluating transitions across a session
engine = GovernanceEngine("model.xsd")
result = engine.evaluate_transition(current_state="draft", target_state="review")
print(result.decision)
print(result.receipt.receipt_hash)   # SHA-256 tamper-evident receipt

git clone https://github.com/SemanticDataCharter/SDCRM.git
cd SDCRM
# See sdc4/schemas/sdc4.xsd (source of truth)
# See sdc4/specification/sdc4-specification.md

We welcome contributions to all SDC projects!

• Code - Implement features, fix bugs, improve performance
• Documentation - Write guides, improve examples, translate content
• Testing - Write tests, report bugs, verify fixes
• Use Cases - Share how you're using SDC
• Feedback - Suggest improvements, discuss architecture

1. Read the Contributing Guidelines
2. Agree to the Contributor License Agreement before your first merge. You keep ownership of your work; see the CLA README for what it grants and why.
3. Check the Code of Conduct
4. Review the Security Policy
5. Find a project that interests you
6. Open an issue or PR

Each repository has its own CONTRIBUTING.md and CLAUDE.md files with specific guidance for that project.

We take security seriously. See our Security Policy for:

• Vulnerability reporting
• Security best practices
• Response timeline

Report vulnerabilities: security@axius-sdc.com

• Website: https://semanticdatacharter.com
• Specification: sdc4-specification.md
• Practitioners: https://axius-sdc.com/practitioners/
• PyPI Packages: sdcvalidator | sdcgovernance | sdcreceipt | form2sdc
• AI Instructions: ai.txt

Controlled and maintained by: Axius SDC, Inc. Contact: contact@axius-sdc.com License: Apache 2.0 (specification and open source tools)

Founders:

• Timothy W. Cook - Founder & CEO
• Dr. Luciana Tricai Cavalini, PhD - Co-Founder & Partner
• Dr. Nikki Shaw, PhD - Co-Founder & Partner

Team: International (US, Canada, Brazil), 40+ years combined experience

• Website: semanticdatacharter.com
• Company: axius-sdc.com
• LinkedIn: Axius SDC
• Email: contact@axius-sdc.com

• SDC Specification: Apache 2.0
• Open Source Libraries: Apache 2.0
• Trademarks: "Semantic Data Charter" and "SDC" are trademarks of Axius SDC, Inc.

The autonomous era needs deterministic data. The substrate is ready.